世界许多地区的Covid-19疫苗的快速推出是大流行的最大成功故事，可防止病毒病毒的严重疾病，死亡和长期副作用，并为经济复苏做出了努力。但是，新型高度突变的Omicron变体的出现引起了人们对当前疫苗作物不再足够有效的担心。

COVID-19，SARS-COV-2背后的病毒与其他更具致命的冠状病毒（如SARS-COV-1和中东呼吸道综合征（MERS））密切相关。我们相对幸运的是，这是造成大流行病的家庭的危险较小（尽管更具传染性）。使这些病毒能够开始感染和传播人类之间的事件一直在发生，这只是时间问题，直到再次发生。

尽管仍然有一条漫长的路来控制19009年的大流行 - 而世界则焦急地等待有关Omicron的潜在影响的答案，但可以肯定的是：我们需要做更多的事情才能保持领先地位。这是我们现在看到的病原体的变体形式，还是下一个大流行。

抗原设计是更有效疫苗的关键

对疫苗科学技术开发的投资可以为我们提供对具有流行潜力的病原体反应的工具。

In recent years, vaccine research has largely centered on the development of better and safer delivery systems, such as viral vectors, DNA and mRNA-based vaccines, as well as more effective immune-stimulating adjuvants. However, far less attention has been paid to the component at the heart of every vaccine: the antigen. An effective vaccine relies on effectively presenting information about a pathogen to the immune system, training it to respond should we encounter the pathogen for real. Even if a vaccine has a highly effective delivery system and adjuvant, it is worthless if the body's immune cells and antibodies cannot recognize the pathogen that circulates and causes disease.

此外，如果病原体演变并且蛋白质的变化已被选为疫苗抗原，那么它可能能够逃避疫苗介导的保护。我们在流感的情况下看到了这一点，该疫苗会迅速突变，并要求每年设计新的疫苗。还有许多其他病原体，我们根本没有找到合适的抗原，因此没有疫苗。

Drawing again on the example of COVID-19, the currently available vaccines are all based on the Spike protein that decorates the surface of the virus. This is an ideal antigen, as antibodies can neutralize the virus before it enters the cells. Focusing on the Spike alone meant that vaccine developers were able to rapidly create a highly effective vaccine that could be rolled out worldwide in less than a year. However, significant mutations in the Spike could render these vaccines less potent, or even useless.

创建通用疫苗

解决方案是开发具有更广泛保护的通用，防止未来的疫苗，这意味着它们不仅可以保护现在存在的变体，而且可以保护那些可能出现的变体，以及与现在存在的变体。

实现这一目标的主要瓶颈是鉴定捕获电流和新兴病原体的抗原。这不是一项简单的任务，而是新技术，包括基因组学，蛋白质组学和数据科学的进步，现在为如何设计抗原为通用，防止未来的疫苗设计抗原提供了前所未有的见解。

许多学术界，初创企业和生物技术的研究人员都在解决这个问题，利用大数据和机器学习来了解病原体的进化以及如何最好地代表疫苗抗原的多样性。

One approach developed by theoretical biologist Bette Korber looks for key variable regions in different strains or species, using this information to create a mosaic antigen protein. This is目前正在针对艾滋病毒进行测试，对于疫苗开发人员来说，最大的挑战之一。其他想法依赖于确定在不同菌株甚至物种之间保存和共享的区域。

Sifting through all this biological data in search of insights about the most appropriate antigens is a task well suited to artificial intelligence (AI). AI algorithms can analyze large and varied genomic, evolutionary, clinical and epidemiological datasets to find critical pieces of proteins in pathogens that either highly variable or highly conserved.

This information can then be used to create entirely artificial “pick and mix” antigens formed by a string of epitopes (the units within a protein recognized by antibodies), providing the immune system with everything it needs to know to mount an effective response, both now and in the future. AI-informed design has the potential to create universal, future-proof vaccines that will be effective across a family of closely related pathogens. This concept has already underpinned the development of anovel vaccine for the mosquito-borne tropical disease chikungunya, and atherapeutic vaccine for treating Human Papillomavirus (HPV)这会导致宫颈和其他癌症。这两种疫苗都在临床试验中，而其他疫苗也在进行人类和动物疾病（包括Covid-19，疟疾和非洲猪发烧）中。

我们可能总是被传染病锁定在进化的军备竞赛中，但我们身边有科学。使用更智能的抗原设计来开发通用的，未来的疫苗将使我们能够领先一步，帮助保护世界免受新兴威胁的影响，并减少在这种规模上再次发生的另一个大流行的机会。